Fundamentals

What Defines a Strategy?

A trading strategy is a Python function named strategy that takes stock data as input and returns the same data with a column of trading signals.

All strategies, regardless of complexity, follow this structure:

def strategy(data):

    # Define your strategy logic in here! #

    return data

As long as this structure is in place, you can add any logic or helper functions needed to generate trading signals.

Caution

You must define a function named strategy, or your strategy will be invalid and won’t run. Always start with this basic skeleton structure!

Where to Place the Trading Signals?

The returned data must include a column called signal (case insensitive) with trading signals. You can (and should) add additional columns for analysis, but the signal column is essential.

Caution

If your table lacks a column named signal, your strategy will fail. Our software specifically looks for this column!

Keep reading this documentation to learn more about creating trading signals.

Using the Strategy

Once defined, your strategy is not tied to a specific stock or date range. It can be applied to any stock, time range, or backtesting setup you choose.

All trades are simulated to execute at the closing price to eliminate feedforward bias and ensure that indicators like the high and low of the period remain valid for decision-making. We considered making this configurable, but enforcing execution at close provides a clear and realistic benchmark while preventing lookahead issues in backtesting.